I'EBS 21183
FEBS Letters 440 (1998) 59 63
Demonstration of thermal dissipation of absorbed quanta during
energy-dependent quenching of chlorophyll fluorescence in
photosynthetic membranes
Wafaa Yahyaoui, Johanne Harnois, Robert Carpentier*
Groupe de Recherche en Enec~ie et h~/brmation Biomoh;culaires. University; ~h~ Qt,;hec ~'t Trois-Rivi~res. P.O. Bo.v 500. Trois-Rivibres,
Que. GgA 5H7, Cana~ht
Received 5 October 1998
lbstract When plant leaves or chloroplasts are exposed to
illumination that exceeds their photosynthetic capacity, photoprotective mechanisms such as described by the energy-dependent
non-photochemical) quenching of chlorophyll fluorescence are
involved. The protective action is attributed to an increased rate
constant for thermal dissipation of absorbed quanta. We applied
photoacoustic spectroscopy to monitor thermal dissipation in
spinach thylakoid membranes together with simultaneous measurement of chlorophyll fluorescence in the presence of inhibitors
of opposite action on the formation of ApH across the thylakoid
membrane (tentoxin and nigericin/valinomycin). A linear relationship between the appearance of fluorescence quenching
during formation of the ApH and the reciprocal variation of
thermal dissipation was demonstrated. Dicyciohexylcarbodiimide, which is known to prevent protonation of the minor
light-harvesting complexes of photosystem II, significantly
reduced the formation of fluorescence quenching and the
concurrent increase in thermal dissipation. However, the addition
of exogenous ascorbate to activate the xanthophyll de-epoxidase
increased non-photochemical fluorescence quenching without
affecting the measured thermal dissipation. It is concluded that
a portion of energy-dependent fluorescence quenching that is
independent of de-epoxidase activity can be readily measured by
photoacoustic spectroscopy as an increase in thermal deactivation processes.
© 1998 Federation of European Biochemical Societies.
Key n'ords." Non-photochemical fluorescence quenching;
Photoacoustic spectroscopy; Thylakoid membrane:
Photosynthesis, Proton gradient; Xanthophyll
1. Introduction
Exposure of plant leaves or chloroplasts to illumination
that exceeds their photosynthetic capacity leads to photoinhibition and degradation of the electron transport system. To
avoid premature photoinhibition, excessive excitation energy
can be dissipated by protective mechanisms such as described
by the energy-dependent (non-photochemical) quenching of
chlorophyll fluorescence [1,2]. It is thought that chlorophyll
fluorescence is quenched by zeaxanthin (and antheraxanthin)
molecules bound to the minor chlorophyll binding proteins of
photosystem II [3,4]. Building of the pH gradient, presumably
in the localized membrane domains [5], would activate the
violaxanthin de-epoxidase with the resulting formation of antheraxanthin and zeaxanthin molecules [6] and promote the
formation of binding sites on the chlorophyll binding proteins
*Corresponding author. Fax: (1) (819) 3765057.
for these xanthophylls [7]. Protonation of the proteins would
be a prerequisite for xanthophyll binding [4]. It is proposed
that de-epoxidized xanthophylls could directly trap the singlet
excited states of chlorophyll a whereas violaxanthin cannot
[8]. Alternatively, it is suggested that the binding of zeaxanthin or protonation itself would favor an aggregated state of
the chlorophyll binding proteins with the concurrent formation of weakly fluorescent traps for chlorophyll excited states
[9]. Another hypothesis implies that the increased pH in the
thylakoid lumen deactivates the oxygen evolving complex of
photosystem I1 resulting in the formation of the oxidized primary donor P680 + which would directly quench chlorophyll
fluorescence [10].
The protective action of energy-dependent fluorescence
quenching against excess light is attributed to an increased
rate constant for radiationless energy dissipation processes
in the antenna chlorophyll as calculated from chlorophyll fluorescence data [3]. Thermal dissipation processes can be measured by photoacoustic spectroscopy. It was recently shown,
using this technique, that photochemical quenching of chlorophyl1 fluorescence due to open photosystem II reaction centers
is directly correlated with a coinciding photochemical quenching of thermal dissipation [11 13]. An increased thermal dissipation during non-photochemical fluorescence quenching
was also reported using photoacoustic spectroscopy [14-17],
though some other reports concluded that energy-dependent
quenching was not associated with significant modification in
the yield of thermal dissipation [18,19].
In this report, we use photoacoustic spectroscopy to monitor thermal dissipation simultaneously with chlorophyll fluorescence in spinach thylakoid membranes under conditions
where energy-dependent fluorescence quenching is observed.
Various inhibitors were used to support the clear demonstration that energy-dependent fluorescence quenching is directly
correlated with an increase in thermal deactivation processes.
2. Materials and methods
Thylakoid membranes were isolated from deveined spinach leaves
as reported previously [20] and resuspended at 2 mg chlorophyll/ml in
330 mM sorbitol, 50 mM TES-NaOH (pH 7.5) and 2 mM MgC12.
When required, the inhibitors were prepared in 80% acetone and less
than 1% of the stock solution was mixed with the thylakoid preparations. Thylakoid membranes (250 p,g chlorophyll/ml, 1 ml) were aspirated onto a nitrocellulose filter (Millipore, 0.4 pM pore size) using a
gentle vacuum to obtain a homogeneous layer of photosynthetic
membranes [21]. Simultaneous fluorescence and thermal dissipation
measurements were performed with a laboratory-constructed system
using a photoacoustic cell (MTEC Photoacoustic Inc., Ames, IA) in
combination with a PAM-101 chlorophyll fluorometer (Walz, Effeltrich, Germany J, as previously described [11,13].
0014-5793/98/$19.00 4' 1998 Federation of European Biochemical Societies. All rights reserved.
Pll: S00 1 4 - 5 7 9 3 ( 9 8 ) 0 [ 430-6
W. Yahyaoui et al./FEBS Letters 440 (1998) 59~53
60
0.8
"
Fm I
Qml
~
"
Qm
0.6
0.4
0.2
Fo
3
I
r1.3
0
O
•<
O
O
Qt
D
A
1
11
ITT2
211
1
0.8
0.60.4-
O
0.2
E
Z
0
11
11
1~"T2
11
ffTz
2fil
1
0.8
O
0.6-
0.2
0
i
C
11
F
2111
Fig. 1. Simultaneous recordings of fluorescence (left panels) and thermal dissipation (right panels) in thylakoid membranes. A and D: With the
addition of 0.1 gM tentoxine. B and E: Control. C and F: With the addition of 5 gM nigericin and valinomycin. Up and down arrows indicate light on and off, respectively: light 1, weak 1.6 kHz fluorometer probe beam; light 2, intensity-modulated photoacoustic measuring beam
(680 nm, 3.8 Wlm 2, 35 Hz). Pulses of saturating white light are applied for 2 s each at 5 s intervals.
3. Results and discussion
Typical traces obtained from simultaneous recordings of
chlorophyll fluorescence (pulse amplitude modulated fluorometer) and thermal dissipation (photoacoustic signal) in spinach
thylakoid membranes are presented in Fig. 1. In these measurements, the weak 1.6 kHz fluorometer probe beam induces
the fluorescence rise to the initial level Fo. The analogous
thermal dissipation level Qo is not detectable directly as explained in a previous report [11]. The photoacoustic signal Qt
is induced by an intensity-modulated measuring beam (680
nm, 3.8 W/m 2, 35 Hz). Upon absorption of this modulated
light, thermal deactivation processes produce a pressure wave
at the surface of the sample which is detected by a sensitive
microphone connected to a lock-in amplifier. The modulated
photoacoustic measuring beam also raises the fluorescence to
an intermediate level of variable fluorescence Ft due to electron transport in photosystem II.
The total closure of the reaction centers is obtained by the
further addition of saturating pulses of non-modulated light.
This will provide the conditions for the specific measurement
of non-photochemical quenching in photosystems where the
photochemical activity is saturated. The 2 s pulse duration is
required to provide an accurate acoustic signal measurement.
The relatively weak intensity of the photoacoustic measuring
beam and the low duration of saturating pulses together with
the very short duration of an entire experiment (128 s) prevent
any photoinhibition. The first pulse provides the maximal
fluorescence level FI,~ and a thermal dissipation level Q./.
During illumination with the photoacoustic measuring beam
and the saturating pulses, a pH gradient is progressively
formed across the thylakoid membranes and the energy-dependent fluorescence quenching gradually increases. The
above causes the progressive decline of F,,, to various F,,,'
levels. It is shown in Fig. 1 that this decline in chlorophyll
fluorescence coincides with a gradual increase of the thermal
Q]II' to the level Qm that determines the maximal
thermal dissipation obtained at the last pulse used.
Various terminology and equations have been used to
quantify non-photochemical fluorescence quenching (NPQ)
or the quenching coeffÉcient q~ [22]. However, the increasing
thermal dissipation from Qm' to Qln observed during the formation of the pH gradient across the thyla'koid membrane
(Fig. 1) is better quantified as (Q,.-QI.')/Qm × 100%, a parameter that can be viewed as a coefficient of non-photochemical quenching of thermal dissipation, qgQ. To use a term
analogous to what can be used for thermal dissipation, nonphotochemical fluorescence quenching will be expressed as
qNF= (F,,~-F,,, ')IF,, x 100%.
The values obtained at each saturating pulse for both qNQ
dissipation
61
i< Yahyaoui et al./FEBS Letters 440 (1998) 5963
~nd qNF are quantified in Fig. 2 (circles) together with the
alues obtained in the presence of two inhibitors of opposite
~,ction tentoxin and nigericin/valinomycin. Fig. 2 shows that
,~~F obtained from a control experiment such as shown in Fig.
! increases with illumination period. Also, during the increase
f qxF, qNQ is progressively released (Fig. 2) as calculated
l'om the increasing Q,f values seen in Fig. 1. The quenching
, alues are kept relatively small due to the short duration of
~e saturating illumination as mentioned above. Tentoxin, an
ncompetitive inhibitor of the chloroplast CF1-ATPase [6], is
nown to induce an increase in ApH [23] and consequently an
1crease in energy-dependent fluorescence quenching, in comarison with the control experiments, is found in Fig. 2. Conersely, the couple nigericin/valinomycin, which is known to
, issipate the electrochemical H + gradient, strongly reduced
qe fluorescence quenching. The corresponding variations of
, xQ are also seen in Fig. 2. The thermal deactivation procsses increased to an higher extent (qxQ decreases) in the
resence of tentoxin in comparison with the control experilent which is apparent from a higher initial value of qNQ.
: :urther, the couple nigericin/valinomycinreduced the gradual
12
10
8
;~Z 6
4
2
0
I
I
I
I
I
20
24
1.6
~.
U
1.2
z 0.8
0.4
0
0
4
8
12
16
PULSES
28
:ig. 2. Values for qNF and qNQ obtained from experiments such as
~resented in Fig. 1 as a function of the saturating pulse number: o,
vithout additive: A, with 0.1 pM tentoxin; I , with 5 pM nigericin
md valinomycin: ©, with 0.5 mM dicyclohexylcarbodiimide. The
:alues for F,,,' and Q,,,' are obtained from averaging the values
neasured during a given saturating pulse.
°f
0.0
0
2
4
6
8
10
12
qN F (%)
Fig. 3. Linear correlation between qxF and q~Q (the correlation
factor R is given in parentheses) obtained from the data of Fig. 2:
O, without additive (R=0.988): i , with 0.1 BM tentoxin (R=
0.994): I, with 5 BM nigericin and valinomycin (R=0.980).
increase in thermal deactivation coinciding with a weaker initial value of qxQ. In these experiments, the inhibitors elicited
only a partial effect in comparison with the expected action of
these compounds at the concentrations used. However, the
samples are aspirated onto a nitrocellulose filter before measurements. This procedure dramatically increases the chlorophyll/inhibitor ratio.
In control experiments, an increase of q\.F by 7.5% corresponded to a decrease of qNQ by 1.3% indicating that total
fluorescence emission with the photosystems in the closed
state accounted for about 14% of the de-excitations processes
which nearly corresponds to the value of 10% previously suggested [24]. The values of q_~Q and qxF were affected in similar proportion by the uncouplers (about 40% increase or
decrease of maximal values, respectively, at the uncoupler
concentrations used). Further, in all cases studied in Fig. 2,
a clear linear relationship, with correlation factors near unity,
was found between qxF and qNQ (Fig. 3). The above demonstrates that the changes in fluorescence intensity are strictly
correlated with the increase in thermal dissipation and therefore that the energy-dependent fluorescence quenching is associated with a proportional increase in thermal deactivation
processes.
As mentioned, protonation of pigment binding proteins
may be a prerequisite for non-photochemical quenching of
chlorophyll fluorescence. In line with the above, we have
used the known inhibitor of energy-dependent fluorescence
quenching dicyclohexylcarbodiimide (DCCD). It was shown
that DCCD short circuits the proton pumping activity of
photosystem II due to its binding to the minor pigment binding proteins and thus prevents protonation at these sites [25].
Interestingly, the inhibition of qxF by DCCD is accompanied
by the corresponding effect on qxQ demonstrating that protonation of minor pigment binding proteins is involved in the
increased thermal dissipation observed during fluorescence
quenching.
It is demonstrated that activation of the enzyme violaxanthin de-epoxidase requires both low pH and ascorbate [26]. In
isolated thylakoid preparations, endogenous ascorbate is re-
62
W. Yahyaoui et al./FEBS Letters 440 (1998) 5943
20
15
3
A
Z
Z
10
2 ~
1
0
30
35
40
ASCORBATE
50
60
(mM)
Fig. 4. Maximal values of q~F and qNQ obtained from data such as presented in Fig. 2 but with various ascorbate concentrations. The maximal values were obtained at the last pulse applied (25th) for qNF and at the first pulse for qNQ. Inset: Values for qNF obtained from experiments such as presented in Fig. 1 as a function of the saturating pulse number: ©, without additive; e, with 60 mM ascorbate. The values of
F,, and Qlll were normalized to 1 before calculation.
moved during isolation procedures and the fluorescence
quenching measured is likely independently from de-epoxidase
activity. Ascorbate can be added to isolated thylakoid membranes or to the purified enzyme to accelerate the violaxanthin
cycle [26]. As expected, addition of ascorbate to the thylakoid
preparations resulted in a substantial increase in qNF (Fig. 4,
inset) showing that the enhanced formation of de-epoxidized
xanthophylls increased non-photochemical fluorescence
quenching during illumination. This increase was dependent
on ascorbate concentration (Fig. 4). Interestingly, qxQ remained totally unaffected by exogenous ascorbate at all the
ascorbate concentrations used (Fig. 4).
The apparent absence of an increased thermal dissipation
following the addition of ascorbate indicates that the enhanced thermal dissipation is not measured by the technique
used. Indeed, only the events with a life-time short enough to
follow the modulation frequency of the incident photoacoustic
measuring beam are monitored by the corresponding pressure
wave. It has been suggested that energy dissipation through a
retarded pathway such as energy transfer from chlorophylls to
carotenoids and thermal dissipation of the carotenoid triplet
states may be responsible for a sufficient delay that would
impair the detection by the modulated technique [27]. This
is consistent with the lack of correlation between the increased
thermal dissipation measured by photoacoustic spectroscopy
and zeaxanthin formation during non-photochemical fluorescence quenching in leaves [16].
The present demonstration of the enhanced thermal dissi-
pation of absorbed quanta in relation to the energy-dependent
fluorescence quenching together with the modulation-independent positive action of exogenous ascorbate in the nonphotochemical quenching coincides with the current mechanistic view of this photoprotection mechanism regarding the
involvement of two different processes, one that depends on
the de-epoxidase activity and the other that is independent of
it. In the former process, fluorescence quenching may occur in
the pigment binding proteins protonated upon formation of
the localized pH gradient. In that case, newly de-epoxidized
xanthophyll molecules could readily bind to their site to
quench the excited states of chlorophyll [2,7]. Alternatively,
the fluorescence quenching may proceeds through de-epoxidase-independent processes. Then, protonation of the pigment
binding proteins could provide the conformational changes
required for the formation of dissipating centers as previously
proposed [9] or inhibition of oxygen evolution and formation
of dissipating photosystem II reaction centers may be involved
[10]. All these de-epoxidase-independent processes are likely
to proceed at the modulation frequency of the incident photoacoustic measuring beam because they are directly related to
the formation of the ApH. Our results showing the inhibition
of qxQ in the presence of DCCD demonstrate that the protonation of pigment binding proteins is implicated in the enhanced thermal dissipation. Those data favor the hypothesis
of thermal dissipation in the pigment bead.
Acknowh,dgemenls." The authors thank A. Tessier for his professional
I i Yahyaoui et al./FEBS Letters 440 (1998) 59~63
t~,sistance. This work was supported by the Natural Sciences and
1 ngineering Research Council of Canada.
keferences
l] Detaining-Adams, B. (1990) Biochim. Biophys. Acta 1020, l 24.
2] Gilmore, A.M. (1997) Physiol. Plant. 99, 197 209.
3] Walters, R.G., Ruban, A.V. and Horton, P. (19941 Eur. J. Biochem. 226, 1063 1069.
4] Pesaresi, P., Sandon',i, D., Giuffra, E. and Bassi, R. (1997) FEBS
Lett. 402, 151 156.
15] Pftindel, E.E., Renganathan, M., Gilmore, A.M., Yamamoto,
H.Y. and Dilley, R.A. (1994) Plant Physiol. 106, 1647 1658.
16] Hu, D., Fiedler, H.R., Golan, T., Edelman, M., Strotmann, H.,
Shavit, N. and Leu, S. (1997) J. Biol. Chem. 272, 5457 5463.
17] Gilmore, A.M., Hazlett, T.L., Debrunner, P.G. and Govindjee
(1996) Photochem. Photobiol. 64, 552-563.
~8] Frank, H.A., Cua, A., Chynwat, V., Young, A., Gosztola, D.
and Wasielewski, M.R. (1994) Photosynth. Res. 41, 389 395.
[9] Ruban, A.V. and Horton, P. (1992) Biochim. Biophys. Acta
1102, 30 38.
0] Bruce, D., Samson. G. and Carpenter, C. (1997) Biochemistry 36,
749 755.
I] Allakhverdiev, S.I., Klimov, V.V. and Carpentier, R. (1994)
Proc. Natl. Acad. Sci. USA 91, 281 285.
2] Markovic, D. and Carpentier, R. (1995) Biochem. Cell Biol. 73.
247 252.
3] Yahyaoui, W., Markovic, D. and Carpentier, R. (1997) Opt. Eng.
36, 337 342.
63
[14] Snel, J.F.H., Kooijman, M. and Vredenberg, W.J. (1990) Photosynth. Res. 25, 259-268.
[15] Dau, H. and Hansen, U. (1990) Photosynth. Res. 25, 269 278.
[16] Havaux, M. and Tardy, F. (1997) J. Photochem. Photobiol. B.
40, 68 75.
[17] Mullinaux, C.W., Ruban, A.V. and Horton, P. (1994) Biochim.
Biophys. Acta 1185, 119 123.
[18] Buschmann, C. and Kocsanyi. L. (1989) Photosynth. Res. 21,
129 136.
[19] Havaux, M. (1990) Photochem. Photobiol. 51, 481 486.
[20] Purcell, M., Leroux, G.D. and Carpentier, R. (1990) Pestic. Biochem. Physiol. 37, 83 89.
[21] Velitchkova, M. and Carpentier, R. (1994) Photosynth. Res. 40,
263 268.
[22] Demmings-Adams, B., Adams Ill, W.W., Baker, D.H., Logan,
B.A., Bowling, D.R. and Verhoeven, A.S. (1996) Physiol. Plant.
98, 253 264.
[23] Quick, P., Sheibe, R. and Stitt, A. (1989) Biochim. Biophys. Acta
974, 282 288.
[24] Lavorel, J. and Etienne, A.L. (19771 in: Primary Processes in
Photosynthesis (Barber, J., Ed.), pp. 203 268, Elsevier/North
Holland, Amsterdam.
[25] Jahns, P. and Junge, W. (199(/) Eur. J. Biochem. 193, 731 736.
[26] Bratt, C.E., Arvidsson, P.-O., Carlsson. M. and Akerlund, H.-E.
(1995) Photosynth. Res. 45, 169 175.
[27] Gruszecki, W.I.. Wojtowicz, K., Krupa, Z. and Strzalka, K.
(1994) J. Photochem. Photobiol. B. 22. 23 27.